To assess the validity of tympanometry as a test for the presence of middle ear effusion using a 'gold-standard' of myringotomy performed after a nitrous oxide-free general anaesthetic, we performed a prospective validity study comparing tympanometry traces obtained immediately pre-operatively from patients undergoing grommet insertion, with the otomicroscopic findings at myringotomy. Nitrous oxide was omitted from the anaesthetic gaseous mixture as it has been reported to displace middle ear effusions. One hundred and seventy-two patients (aged 1.5-15 years) with a clinically assessed 3 month history of middle ear effusion were included in the study. Sensitivity and specificity of a Jerger classification Type B tympanometric trace for the presence of middle ear effusion were 0.73 and 0.84, respectively. We conclude that tympanometry is a valid test in assessing the presence of middle ear effusion compared to a 'gold standard' of myringotomy performed after a nitrous oxide-free general anaesthetic.